The scientific goal of this ROO proposal is to investigate polymorphisms in specific candidate genes that may be associated with indices of cognitive, functional, and structural brain integrity as intermediate phenotypes (endophenotypes) of dementia. The assumption of the endophenotype strategy is that gene effects at the level of the brain are more direct than the complex behavior, and will show association in carriers of risk alleles even if the carriers show no clinical diagnostic characteristics.The primary aims are to: (1) measure variation in five candidate genes (APOE, SORL1, BDNF, COMT, and KIBRA), all with known involvement in either dementia or cognitive impairment, and examine the relationships between candidate gene variants and cognitive tasks allowing for cross-species comparisons and probing either frontal (reversal learning, elemental discriminations) or medial temporal (virtual water maze, transverse patterning) lobe function; (2) examine genotype differences in brain activaty patterns (fMRI; Default Mode Network), and (3) examine possible mechanisms underlying the gene-behavior relationship through morphological (MRI voumetrics and DTI) and biochemical (IH MRSI) endophentypes, employing a multi-modal MR imaging approach. This will be accomplished through the analyses of DNA , cognitive, and multi-modal imaging data collected from approximately 150 middle-aged, non-demented, healthy adults without family history of dementia. Proposed experiments will provide converging evidence on several different levels of investigation that can also be cross-compared to the findings from animal literature. Insights will be gained into cognitive deficits and brain changes in middle-aged adults prior to any manifestation of clinical symptoms, which have been relatively understudied in comparison to both younger and older adults. The findings will lead to a better understanding of pathogenic loci and target pathways for therapy, which could result in possible additional treatment strategies. These studies will provide a genetic, biological, clinical and behavioral background for future ROI applications aimed at identifying new genes associated with age-related cognitive decline and dementia.